SYNTHESIS AND APPLICATION OF MOLECULARLY IMPRINTED SOLID-PHASE EXTRACTION FOR THE DETERMINATION OF TERBUTALINE IN BIOLOGICAL MATRICES LIM LAY LEE UNIVERSITI SAINS MALAYSIA 2006
SYNTHESIS AND APPLICATION OF MOLECULARLY IMPRINTED SOLID-PHASE EXTRACTION FOR THE
DETERMINATION OF TERBUTALINE IN BIOLOGICAL MATRICES
LIM LAY LEE
UNIVERSITI SAINS MALAYSIA
2006
SYNTHESIS AND APPLICATION OF MOLECULARLY IMPRINTED SOLID-PHASE EXTRACTION FOR THE DETERMINATION OF TERBUTALINE IN
BIOLOGICAL MATRICES
by
LIM LAY LEE
Thesis submitted in fulfillment of the
requirements for the degree of Master of Science
DECEMBER 2006
ii
ACKNOWLEDGEMENTS
With a deep sense of gratitude, I wish to express my sincere thanks to my
supervisor, Associate Professor Dr. Tan Soo Choon, for his immense guidance,
dedication and motivation throughout the project years. His wide knowledge, valuable
advice and encouragement have been of great value for me and will be remembered
lifelong. It is a pleasure to have the opportunity to learn and work with him.
Besides that, I will like to extend my warmest thanks to those whom have
helped throughout my work especially to the staffs of Veterinary Forensic Laboratory.
All their kind assistance and support especially by Miss June Sim, Miss Lye Jin Sean
and Mr. Ng Chek Wan are greatly appreciated. It will be impossible to complete my
research as scheduled without their kind arrangement especially regarding to the
usage of GC-MS instruments. The financial support of Universiti Sains Malaysia
through the Post Graduate Teaching Scheme is gratefully acknowledged.
Last but not least, I am grateful to my family and friends for their
encouragement, moral support and understanding. Finally, I will like to take this
opportunity to thank all those whom I have not mentioned. Their direct and indirect
assistance have had an impact to this thesis.
The whole experience of pursuing this MSc programme has taught me to be
more independent, patient and in understanding the true meaning of “research”.
iii
TABLE OF CONTENTS
Page ACKNOWLEDGEMENTS ii
TABLE OF CONTENTS iii
LIST OF TABLES viii
LIST OF FIGURES xii
LIST OF PLATES xix
LIST OF ABBREVIATION xx
LIST OF SYMBOLS xxiii
ABSTRAK xxiv
ABSTRACT xxvi
CHAPTER ONE: INTRODUCTION TO MOLECULARLY IMPRINTED POLYMER (MIP)
1.1 Introduction 1
1.2 General Principle of Molecular Imprinting 2
1.3 Molecular Imprinting Approaches 4
1.3.1. Covalent Imprinting
1.3.2. Non-covalent Imprinting
1.3.3. Comparison Between Covalent and Non-covalent
Imprinting
4
5
6
1.4 Synthesis of Molecularly Imprinted Polymer 7
1.4.1. Polymerisation Method
1.4.1.1. Bulk Polymerisation
1.4.1.2. Suspension Polymerisation
1.4.1.3. Precipitation Polymerisation
1.4.1.4. Two-step Swelling Polymerisation
1.4.1.5. Emulsion Core-shell Polymerisation
1.4.2. Polymerisation Reagents
1.4.2.1. Template Molecule
1.4.2.2. Functional Monomer
1.4.2.3. Cross-linking Agent
1.4.2.4. Porogen
1.4.2.5. Initiator
9
9
9
10
10
11
12
13
14
14
15
15
iv
1.4.3. Factors to Consider in the Synthesis of Selective MIP
1.4.3.1. Molar Ratio of Template: Monomer: Cross-linker
(T: M: X)
1.4.3.2. Stability of Monomer-template Assemblies
1.4.3.3. Polymerisation Temperature
16
16
17
19
1.5 Polymer Structure and Morphology
1.5.1. Types of Pores
1.5.2. Adsorption / Binding Isotherms
20
20
21
1.6 Application of Molecularly Imprinted Polymers 23
1.6.1. Affinity Based Solid-phase Extraction
1.6.2. Sensors and Membranes
1.6.3. Others
23
25
25
1.7 Aims and Objectives of the Present Investigation 26
CHAPTER TWO: PRODUCTION OF SALBUTAMOL IMPRINTED POLY(METHACRYLIC ACID-CO-ETHYLENE GLYCOL DIMETHACRYLATE) AND NON-IMPRINTED POLYMER
2.1 Introduction
2.1.1. Proton Nuclear Magnetic Resonance Spectroscopy (1H
NMR)
2.1.2. Infrared Spectroscopy (IR)
28
29
30
2.2 Aim of Experiment 30
2.3 Materials 31
2.4 Methods
2.4.1. Stock and Working Standard Solutions
2.4.2. Preparation of Chemical Solutions
2.4.3. Extraction of Salbutamol Free Base from Salbutamol
Sulphate Aqueous Solution
2.4.4. Preparation of Salbutamol Imprinted Poly(methacrylic
acid-co-ethylene glycol dimethacrylate) and Non-imprinted
Polymer
2.4.5. Soxhlet Extraction of Salbutamol Template from the
Imprinted Poly(methacrylic acid-co-ethylene glycol
dimethacrylate)
32
32
33
34
34
36
v
2.4.6. Preparation of Molecularly Imprinted Polymer as Solid-
phase Extraction (MIPSPE) Column
2.4.7. Selectivity Test of Imprinted and Non-imprinted Polymers
37
37
2.4.8. Preparation of Samples for 1H NMR Analysis in
Deuterated CD3CN-CD3OD (75: 25 % v/v)
2.4.9. Preparation of Samples for FTIR Analysis
2.4.10. Instrumentation
38
39
39
2.5 Results and Discussion
2.5.1. Synthesis of Imprinted and Non-imprinted Polymers
2.5.2. Removal of Template Molecule from Imprinted Polymer
2.5.3. Selectivity Test of Imprinted and Non-imprinted Polymers
2.5.4. Spectroscopic Analysis of Pre-organised Monomer-
template Assemblies
2.5.4.1. Infrared Spectroscopy (FTIR)
2.5.4.2. Proton Nuclear Magnetic Resonance
Spectroscopy (1H NMR)
40
40
45
46
52
52
55
2.6 Conclusion 61
CHAPTER THREE: EVALUATION OF MIP FOR THE APPLICATION OF SOLID-PHASE EXTRACTION AND MATRIX INTERFERENCE STUDY
3.1 Introduction
3.1.1. General Overview of Solid-phase Extraction
3.1.2. Types of Sorbents and Modes of Interaction
3.1.3. Limitations of Conventional SPE Method
62
62
65
66
3.2 Aim of Experiment 67
3.3 Materials 67
3.4 Methods
3.4.1. Stock and Working Standard Solutions
3.4.2. Preparation of Chemical Solutions
3.4.3. Binding Study
3.4.4. Elution Study
3.4.5. Sample Application / Loading Study
3.4.5.1. Stage 1
3.4.5.2. Stage 2: Method Optimisation
68
68
69
70
72
73
73
74
vi
3.4.6. Matrix Interference Study
3.4.6.1. Dilution with 100 mM Ammonium Acetate pH 7
3.4.6.2. Dilution with 100 mM Ammonium Acetate pH 9
3.4.7. Cross-specificity Study
3.4.8. Instrumentation
75
76
77
77
78
3.5 Results and Discussion 78
3.5.1. Gas Chromatography-Mass Spectrometry Analysis of
Terbutaline and Salbutamol
3.5.2. Binding Study
3.5.3. Elution Study
3.5.4. Sample Application / Loading Study
3.5.5. Matrix Interference Study
3.5.6. Cross-specificity Study
78
82
86
92
102
106
3.6 Conclusion 126
CHAPTER FOUR: QUANTITATIVE AND QUALITATIVE ANALYSIS 4.1 Introduction
4.1.1. Bioanalytical Method Validation
4.1.2. Confirmatory Analysis
127
127
129
4.2 Aim of Experiment 131
4.3 Materials 131
4.4 Methods
4.4.1. Stock and Working Standard Solutions
4.4.2. Preparation of Chemical Solutions
4.4.3. Preparation of Spiked Human Plasma and Urine
4.4.4. MIPSPE Extraction Procedure
4.4.5. SPE Mixed Mode Procedure (Strata Phenomenex Screen
C) as Reference Method
4.4.6. Instrumentation
132
132
133
134
135
137
138
4.5 Results and Discussion
4.5.1. Bioanalytical Method Validation (Quantitative Analysis)
4.5.1.1. Specificity / Selectivity Test
4.5.1.2. Calibration Curve and Correlation of
Determination
139
139
139
141
vii
4.5.1.3. Between Day Validation
4.5.1.4. Within Day Validation
143
144
4.5.2. Method Comparison Study: SPE Mixed Mode as
Reference Method
4.5.2.1 Pearson’s Correlation between Proposed
Method (MISPE) and Reference Method (SPE
Mixed Mode)
4.5.2.2 Advantages of the MIPSPE Method
4.5.3 Confirmatory Analysis (Qualitative Study)
146
146
155
157
4.6 Conclusion 164
CHAPTER FIVE: GENERAL CONCLUSIONS
5.1 General Conclusions 165
5.2 Limitations of Developed Method 170
5.3 Recommendations for Future Research 171
BIBLIOGRAPHY 173
APPENDICES
Appendix A: Signal to Noise Report by Chemstation Software 182
Appendix B: Analytical Method Validation 184
viii
LIST OF TABLES Tables Page
1.1 Summary of advantages and disadvantages of MIPs for
analytical applications (Mahony et al., 2005).
2
1.2 Advantages and disadvantages of covalent and non-covalent
imprinting (Komiyama et al., 2003).
7
1.3 Comparison of different polymerisation methods (Mayes et al.,
1997).
8
1.4 Choice of reagents and molar ratios of T: M: X for the syntheses
of MIPs by bulk polymerisation technique.
18
1.5 Summary of studies in which MIPs have been applied to SPE. 24
2.1 Amount of glacial acetic acid added into ACN to produce various
mixtures of ACN- glacial acetic acid.
33
2.2 Fractions collected for analysis in selectivity test from each
individual cartridge.
38
2.3 Summary of the reagents and amount employed for the
production of MIP by bulk polymerisation method.
41
2.4 Percent breakthrough of terbutaline in each fraction collected
during the loading and elution stage. 10 mL of ACN spiked with
100 ng of terbutaline was loaded into the MIP and blank polymer
(n = 4). Elution was performed with 1 mL of ACN containing
increasing amount of acetic acid (HAc). Terbutaline was
detected in the form of terbutaline tris trimethylsily after
derivatization by MSTFA at 70 oC for 30 minutes. (a) Polymer
Batch # 1; (b) Polymer Batch # 2; (c) Polymer Batch # 3 and (d)
Polymer Batch # 4.
48
2.5 Chemical shift of salbutamol protons in salbutamol solution and
pre-polymerisation mixture of CD3CN-CD3OD (75: 25 % v/v).
57
2.6 Chemical shift of protons in methacrylic acid solution of CD3CN-
CD3OD (75: 25 % v/v).
57
3.1 Common sorbents available for SPE application (Thurman et al.,
1998).
66
3.2 Type of solvents applied as loading medium. 71
3.3 Fractions collected for analysis in binding study. 72
3.4 Fractions collected for analysis in elution study. 73
ix
3.5 Fractions collected for analysis in loading study (Stage 1). 74
3.6 Fractions collected for analysis in loading study (Stage 2). 75
3.7 MIPSPE extraction procedure for urine and plasma samples.
Samples were diluted with 50 mM ammonium acetate pH 7.
76
3.8 Fractions collected for analysis in cross-specificity study. 78
3.9 Diagnostic ions and retention times for the analysis of
terbutaline, salbutamol and nalorphine under SIM mode. All the
analytes were derivatised to form TMS derivatives using MSTFA
at 70 oC for 30 minutes. Chromatographic conditions were as
described in Section 2.4.10.
79
3.10 Binding of terbutaline to imprinted and blank polymers in
different solvents (n = 3). Ten mL of spiked solvent at 10 ng/mL
of terbutaline was loaded into each cartridge (total amount
loaded per column was 100 ng).
83
3.11 Amount of terbutaline measured in various stages of SPE
extraction (n = 3). Ten mL of spiked ACN at 10 ng/mL was
loaded into each column. Different eluting solvents were applied
as a) ACN-glacial acetic acid (95: 5 % v/v); b) ACN-glacial acetic
acid (90: 10 % v/v); c) ACN-glacial acetic acid (80: 20 % v/v); d)
ACN-trifluoroacetic acid (99: 1 % v/v) and e) ACN-triethylamine
(99: 1 % v/v).
87
3.12 Amount of terbutaline measured in various stages of SPE
extraction (n = 3). Five mL of spiked solution at 20 ng/mL
terbutaline was loaded into each cartridge. Different loading
solutions (Stage 1) were applied as a) Distilled water; b) 50 mM
ammonium acetate at pH 5; c) 50 mM ammonium acetate at pH
7 and d) 50 mM ammonium acetate at pH 9.
95
3.13 Amount of terbutaline measured in various stages of SPE
extraction (n = 3). Five mL of spiked solution at 20 ng/mL of
terbutaline was loaded into each cartridge. Different loading
solutions (Stage 2) were applied as a) 50 mM ammonium
acetate at pH 7; b) 50 mM ammonium acetate at pH 9 and c) 50
mM ammonium acetate at pH 11.
99
3.14 Recoveries of terbutaline spiked into diluted urine samples at
various concentration levels. Dilution was conducted at ratio 1: 1
103
x
v/v. Extraction procedure was as described in Section 3.4.6.
Urine samples were from human, bovine and equine.
3.15 Recoveries of terbutaline spiked into diluted human plasma
samples at various concentrations. Extraction procedure was as
described in Section 3.4.6.
105
3.16 Optimisation of MIPSPE extraction procedure for the
determination of terbutaline in plasma and urine samples.
106
3.17 Diagnostic ions and the retention times for GC-MS analysis
under SIM mode. All the analytes were derivatised to form TMS
derivatives using MSTFA or enol solution (for clenbuterol and
metoprolol only) at 70 oC for 30 minutes. Chromatographic
conditions were as described in Section 2.4.10.
108
3.18 Percent breakthrough of analytes at various SPE stages from
the imprinted and non-imprinted polymers (n = 4). Loading
solutions were in a) 50 mM ammonium acetate at pH 7 and b)
ACN.
124
3.19 Log P and pKa value for each of the substance evaluated in
cross-specificity test (Hansch et al., 1990 and Moffat et al.,
1986).
126
4.1 Accuracy and precision data for the determination of terbutaline
in spiked plasma samples at the lower limit of quantification level
(LLOQ at 1 ng/mL) (n = 6).
141
4.2 Equation of calibration curves and correlation of determination
(r2) values for six consecutive days of validation.
143
4.3 Precision for between day validation (n = 6) at low, medium and
high QC samples.
143
4.4 Within day validation, accuracy and extraction recoveries of
terbutaline in spiked plasma samples (n = 7) at the
concentrations of a) 2 ng/mL; b) 5 ng/mL and c) 8 ng/mL.
145
4.5 Diagnostic ions and retention times for the analysis of terbutaline
and salbutamol under SIM mode. All the analytes were
derivatised to form t-BDMS derivatives using N-tert-
butyldimethylsilyl-N-methyl trifluoroacetamide – pyridine (2:1 v/v)
at 60 oC for 1 hour. Chromatographic conditions were as
described in Section 4.4.6.
147
xi
4.6 The concentrations of terbutaline measured by MIPSPE and
SPE mixed mode procedures following the extraction from
spiked plasma samples (n = 20).
152
4.6 The concentrations of terbutaline measured by MIPSPE and
SPE mixed mode procedures following the extraction from
spiked plasma samples (n = 20) (continued).
153
4.7 Summary output of Excel spreadsheet used to analyse the
regression line in Figure 4.8.
154
4.8 Comparison between the retention times and maximum
permitted difference in relative abundance expressed in term of
absolute and relative of the diagnostic ions of terbutaline
extracted from spiked urine samples and pure standards.
162
xii
LIST OF FIGURES
Figures Page
1.1 Schematic diagram of molecularly imprinted polymer. I: self
assembly of template and functional monomers (non-covalent
approach); II: polymerisation process; III: extraction of template
and rebinding of analyte (Adapted from: Turiel et al., 2004).
3
1.2 Schematical representation of the synthetic steps in the different
polymerisation procedures (Adapted from: Perez-Moral et al.,
2004).
8
1.3
1.3
Typical reagents for polymerisation.
Typical reagents for polymerisation (continued).
12
13
1.4 Factors affecting the recognition properties of MIPs related to
the monomer-template assemblies (Adapted from: Sellergren,
1999).
19
1.5 Types of binding sites in MIPs (Adapted from: Sellergren,1999). 21
2.1 Chemical structures of terbutaline and salbutamol. 42
2.2 Schematic diagram of imprinting process. Proposed interactions
between template molecule (salbutamol) and the MIP carboxylic
moieties (MAA). Non-covalent imprinting begins with the
selected template complexing with a functional monomer. Cross-
linking locks the monomers into the proper orientation around
the template. Removal of the template affords selective binding
sites for targeted analyte.
44
2.3 Plot of percent breakthrough and cumulative recovery of
terbutaline versus the SPE stages in imprinted and non-
imprinted/ blank polymers.
50
2.4 Example of extracted ion chromatogram obtained from the E5-
E7 fraction after eluting the MIP cartridge with ACN containing
acetic acid. Derivatisation was carried out using MSTFA reagent
at 70 oC for 30 minutes. Terbutaline tris TMS was quantitated at
m/z = 356, salbutamol tris TMS at m/z = 369 (leaching from MIP)
and nalorphine O, O’ bis TMS at m/z = 455 (as internal
standard). Retention times peak 1: terbutaline tris TMS, 8.53
min; 2: salbutamol tris TMS, 8.99 min; 3: nalorphine O, O’ bis
52
TMS, 12.20 min.
xiii
2.5 Infrared spectrums of a) pre-polymerisation mixture; b)
methacrylic acid and c) salbutamol free base (solid).
54
2.6 Protons position in a) salbutamol and b) methacrylic acid 57
2.7 1H NMR spectrum of salbutamol free base in CD3CN-CD3OD
(75: 25 % v/v).
58
2.8 1H NMR spectrum of methacrylic acid in CD3CN-CD3OD (75: 25
% v/v).
59
2.9 1H NMR spectrum of pre-polymerisation mixture in CD3CN-
CD3OD (75: 25 % v/v).
60
3.1 General process of solid phase extraction. Step 1: Conditioning
of sorbent; Step 2: Sample application; Step 3: Interference
elution or washing and Step 4: Analyte elution (Adapted from:
Thurman et al., 1998).
64
3.2 The mass spectrum for terbutaline tris TMS and postulated
fragmentation pathways (Source: NIST mass spectral
database).
80
3.3 The mass spectrum for salbutamol tris TMS and postulated
fragmentation pathways (Source: NIST mass spectral
database).
80
3.4 The mass spectrum for nalorphine O, O’-bis TMS and postulated
fragmentation pathways (Source: NIST mass spectral
database).
81
3.5 Typical calibration curve (from pure standard) prepared for the
quantification of terbutaline as terbutaline tris TMS derivative
(range from 10 to 120 ng).
81
3.6 Typical calibration curve (from pure standard) prepared for the
quantification of terbutaline as terbutaline tris TMS derivative at
lower amount (range from 2.50 to 30 ng).
82
3.7 Plot of cumulative recovery of terbutaline breakthrough versus
the volume of solvent loaded into the SPE cartridge. The loading
solutions were ACN-glacial acetic acid at a) 99: 1 % v/v; b) 98: 2
% v/v and c) 97: 3 % v/v.
85
xiv
3.8 Plot of percentage breakthrough and cumulative recovery of
terbutaline versus the SPE stages. Different eluting solvents
were applied as a) ACN-glacial acetic acid (95: 5 % v/v); b)
ACN-glacial acetic acid (90: 10 % v/v); c) ACN-glacial acetic
acid (80: 20 % v/v); d) ACN-trifluoroacetic acid (99: 1 % v/v) and
e) ACN-triethylamine (99: 1 % v/v).
89
3.9 Plot of percentage recovery of terbutaline eluted out during
elution step versus the different types of eluting solvents.
Terbutaline measured was from fraction E1 to E4.
91
3.10 Plot of percentage breakthrough and cumulative recovery of
terbutaline versus the SPE stages. Different loading solutions
(Stage 1) were applied as a) Distilled water; b) 50 mM
ammonium acetate at pH 5; c) 50 mM ammonium acetate at pH
7 and d) 50 mM ammonium acetate at pH 9.
97
3.11 Plot of percentage breakthrough and cumulative recovery of
terbutaline versus the SPE stages. Different loading solutions
(Stage 2) were applied as a) 50 mM ammonium acetate at pH 7;
b) 50 mM ammonium acetate at pH 9 and c) 50 mM ammonium
acetate at pH 11.
100
3.12 Plot of percentage recovery of terbutaline versus the pH values
of 50 mM ammonium acetate buffers. 4 mL of eluting solvent
was applied into each cartridge.
101
3.13 Chemical structures of substances tested in the cross-specificity
study.
107
3.14 The mass spectrum for clenbuterol N, O-bis TMS and postulated
fragmentation pathways (Source: NIST mass spectral
database).
109
3.15 The mass spectrum for fenoterol tetrakis TMS and postulated
fragmentation pathways (Source: NIST mass spectral
database).
109
3.16 The mass spectrum for isoxsuprine di TMS and postulated
fragmentation pathways (Source: NIST mass spectral
database).
110
3.17 The mass spectrum for metoprolol TMS and postulated
fragmentation pathways (Source: NIST mass spectral
database).
110
xv
3.18 The mass spectrum for ractopamine tris TMS and postulated
fragmentation pathways. Mass spectrum was obtained from the
analysis of pure ractopamine tris TMS standard at 4 ng/μL under
full scan mode.
111
3.19 The mass spectrum for ibuprofen TMS and postulated
fragmentation pathways (Source: NIST mass spectral
database).
111
3.20 The mass spectrum for boldenone O-TMS and postulated
fragmentation pathways (Source: NIST mass spectral
database).
112
3.21 Typical example of the extracted ion chromatogram of fenoterol
tetrakis TMS obtained from the selective wash fraction of MIP in
which ammonium acetate pH 7 was the loading medium.
Retention times peak 1: nalorphine O, O’ bis TMS, 12.16 min;
2: fenoterol tetrakis TMS, 12.25 min.
112
3.22 Typical example of the extracted ion chromatogram of
isoxsuprine di TMS obtained from the selective wash fraction of
MIP in which ammonium acetate pH 7 was the loading medium.
Retention times peak 1: isoxsuprine di TMS, 11.44 min;
2: nalorphine O, O’ bis TMS, 12.16 min.
113
3.23 Typical example of the extracted ion chromatogram of
clenbuterol N, O-bis TMS obtained from the elution (E1+E2)
fraction of MIP in which ammonium acetate pH 7 was the
loading medium. Retention times peak 1: clenbuterol N, O-bis
TMS, 9.52 min; 2: nalorphine O, O’ bis TMS, 12.23 min.
113
3.24 Typical example of the extracted ion chromatogram of
ractopamine tris TMS obtained from the elution (E1+E2) fraction
of MIP in which ammonium acetate pH 7 was the loading
medium. Retention times peak 1: nalorphine O, O’ bis TMS,
12.20 min; 2: ractopamine tris TMS, 12.39 min.
114
3.25 Typical example of the extracted ion chromatogram of
metoprolol TMS obtained from the elution (E1+E2) fraction of
MIP in which ammonium acetate pH 7 was the loading medium.
Retention times peak 1: metoprolol TMS, 9.56 min; 2: nalorphine
O, O’ bis TMS, 12.16 min.
114
xvi
3.26 Typical example of the extracted ion chromatogram of ibuprofen
TMS obtained from the selective wash fraction of MIP in which
ammonium acetate pH 7 was the loading medium. Retention
times peak 1: ibuprofen TMS, 6.37 min; 2: nalorphine O, O’ bis
TMS, 12.16 min.
115
3.27 Typical example of the extracted ion chromatogram of
boldenone O- TMS obtained from the selective wash fraction of
MIP in which ammonium acetate pH 7 was the loading medium.
Retention times peak 1: nalorphine O, O’ bis TMS, 12.16 min; 2:
boldenone O-TMS, 12.52 min.
115
3.28 Typical calibration curves (from pure standard) for the
quantification of a) clenbuterol; b) fenoterol; c) isoxsuprine; d)
metoprolol; e) ractopamine; f) ibuprofen and g) boldenone as
TMS derivative (range from 10 to 120 ng).
116
3.29 Percent breakthrough of analyte at various SPE stages from
imprinted and non-imprinted polymers (n = 4). Loading medium
were as in a) 50 mM ammonium acetate at pH 7 and b) ACN.
Selective washing was conducted by applying 1 mL of ACN-
glacial acetic acid (99: 1 % v/v) followed by 1 mL x 4 of ACN-
glacial acetic acid (90: 10 % v/v) for the elution step.
122
4.1 Extracted ion chromatogram of terbutaline as terbutaline tris
TMS at m/z = 356 from the blank human plasma and spiked
plasma at 1 ng/mL. Blank and spiked samples were extracted
using the proposed MIPSPE method as described in Table 3.16
but at the dilution ratio of 1: 4 v/v with 62.5 mM ammonium
acetate pH 7 as sample diluent.
140
4.2 Typical calibration curve prepared for the quantification of
terbutaline as terbutaline tris TMS derivative following the
extraction of the drug from spiked plasma.
142
4.3 The mass spectrum for terbutaline tris t-BDMS. Mass spectrum
was obtained from the analysis of pure terbutaline tris t-BDMS
standard at 4 ng/μL under full scan mode.
148
4.4 The mass spectrum for salbutamol tris t-BDMS. Mass spectrum
was obtained from the analysis of pure salbutamol tris t-BDMS
standard at 4 ng/μL under full scan mode.
148
xvii
4.5 The chemical structures of terbutaline and salbutamol as t-
BDMS derivatives and the postulated fragmentation pathways.
149
4.6 Typical example of the extracted ion chromatogram of
terbutaline as t- BDMS derivative following the extraction from
spiked plasma sample at 9 ng/mL. Salbutamol was added as
internal standard. Retention times peak 1: terbutaline tris t-
BDMS, 9.46 min; 2: salbutamol tris t-BDMS , 9.80 min.
149
4.7 Typical calibration curve prepared for the quantification of
terbutaline as terbutaline tris t-BDMS derivative following the
extraction of the drug from spiked plasma by SPE mixed mode
method.
150
4.8 Correlation between terbutaline concentrations in human plasma
(n = 20) measured by the SPE mixed mode method (x axis) and
the proposed MIPSPE method (y axis).
153
4.9 Extracted ion chromatograms at m/z = 356 of pure terbutaline
standard, blank and spiked plasma samples. Extraction
procedures were MIPSPE and SPE mixed mode. Terbutaline
was measured as terbutaline tris TMS derivative by using
MSTFA as the derivatisation reagent for extracts obtained from
both methods.
156
4.10 Full scan mass spectrum of terbutaline tris TMS from spiked
urine sample at 20 ng/mL of terbutaline. Five mL of urine was
diluted at the ratio of 1: 1 v/v before loading into two separate
cartridges. The extracts collected were combined as a single
sample.
158
4.11 Extracted ion chromatogram of terbutaline tris TMS from spiked
urine sample at m/z = 86, 356 and 426. Retention time was at
8.48 min.
158
4.12 Extracted ion chromatogram of terbutaline tris TMS from pure
standard at m/z = 86, 356 and 426. Concentration of the injected
sample was at 1 ng/μL. Retention time was at 8.48 min.
159
4.13 Full scan mass spectrum of terbutaline tris t-BDMS from spiked
urine sample at 20 ng/mL of terbutaline. Five mL of urine was
diluted at the ratio of 1: 1 v/v before loading into two separate
cartridges. The extracts collected were combined as a single
sample.
159
xviii
4.14 Extracted ion chromatogram of terbutaline tris t-BDMS from
spiked urine sample at m/z = 86 and 482. Retention time was at
9.23 min.
160
4.15 Extracted ion chromatogram of terbutaline tris t-BDMS from pure
standard at m/z = 86 and 482. Concentration of the injected
sample was 2 ng/μL. Retention time was at 9.23 min.
160
xix
LIST OF PLATES Plates Page
2.1a Opaque monolith macroporous polymer. 45
2.1b Fine irregular particles within the size of 40 to 75 μm. The
particles were obtained by manually grinding using pestle and
mortar.
45
xx
LIST OF ABBREVIATION
% Percent
% v/v Percent volume per volume
μg/mL Microgramme per millilitre
μL Microlitre
μm Micron, micrometre 1H NMR Proton nuclear magnetic resonance
Å Angstrom
ABDV Azobisdimethylvaleronitrile
ACN Acetonitrile
AIBN Azobisisobutyronitrile
AORC Association of Official Racing Chemist
Blk Blank polymer
C=O Carbonyl group
CD3CN Deuterated acetonitrile
CD3OD Deuterated methanol
CH3COONH4 Ammonium acetate
cm/sec Centimetre per second
cm-1 Wave length in centimetre unit
-COOH Carboxylic acid
CV Coefficient of variation
DVB Divinylbenzene
EC European Council
EGDMA Ethylene glycol dimethacrylate
EIC Extracted ion chromatogram
FDA Food and Drug Administration
FTIR Fourier transform infrared
g Gramme
GC-MS Gas chromatography coupled to mass spectrometry
H2O Water
HAc Glacial acetic acid
HBr Hydrobromide
HCl Hydrochloride
HPLC, LC High performance liquid chromatography
ICH International Conference on Harmonisation
xxi
IOC International Olympic Committee
ISO International Organisation for Standardisation
KBr Potassium bromide
LC-MS Liquid chromatography coupled to mass spectrometry
LLOQ Lower limit of quantification
LOD Limit of detection
Log P Log of octanol-water partition coefficient
M Molar
m/z Mass to charge ratio
MAA Methacrylic acid
MeOH Methanol
mg Milligramme
MIP(s) Molecularly imprinted polymer(s)
MIPCE Capillary electrophoresis based on molecularly imprinted polymer
MIPCEC Capillary electrochromatography based on molecularly imprinted
polymer
MIPSPE Solid-phase extraction based on molecularly imprinted polymer
mM Millimolar
mmHg Millimetre of mercury
MSTFA N-methyl-N-trimethylsilyltrifluoroacetamide
n Number of replicate
NaOH Sodium hydroxide
ng/mL Nanogramme per millilitre
-NH Amino group
NH4I Ammonium iodide
NIP Non-imprinted polymer
NIST National Institute of Standards and Technology oC Degrees celcius
-OH Hydroxyl group
pKa Log of acidity constants
ppm Parts per million
psi Pounds per square inch
QC Quality control
r2 Correlation of determination
rpm Revolutions per minute
Rt Retention time
xxii
SD Standard deviation
SIM Selected ion monitoring
SPE Solid-phase extraction
T: M: X Ratio of template molecule: functional monomer: cross-linker
t-BDMS Tert-butyl dimethylsilyl
TEA Triethylamine
TFA Trifluoroacetic acid
TMS Trimethylsilyl
UV Ultraviolet
xxiii
LIST OF SYMBOLS
δ Chemical shift
Δδ Difference in chemical shift
π Pi
α Alpha
β Beta
∼ Approximately
≥ Equal or greater than
≤ Equal or less than
± Plus minus
xxiv
SINTESIS DAN APLIKASI PENGEKSTRAKAN FASA PEPEJAL BERASASKAN
“MOLECULAR IMPRINTING” UNTUK PENENTUAN TERBUTALINA DI DALAM
SAMPEL BIOLOGI
ABSTRAK
Kajian ini tertumpu kepada penilaian keberkesanan penggunaan MIPs sebagai
bahan padatan dalam SPE untuk pengekstrakan terbutalina di dalam sampel plasma
dan urin. Polimer ini dihasilkan dengan menggunakan asid metakrilik (MAA) sebagai
monomer dan etilena glikol dimetakrilat (EGDMA) sebagai bahan silang manakala
salbutamol dengan struktur yang hampir serupa dengan analit dipilih sebagai molekul
templat. Campuran asetonitril-metanol (75: 25 % v/v) digunakan sebagai pelarut untuk
pempolimeran secara terma pada suhu 50 oC selama 24 jam. Nisbah molar untuk
molekul templat: monomer: bahan silang ialah pada 1: 4: 20. Polimer kawalan tanpa
molekul templat turut dihasilkan melalui teknik yang serupa. Polimer perlu ditumbuk
dan ditapis untuk mendapatkan partikel halus bersaiz 40 hingga 75 μm. Kemudian, ia
diekstrak menggunakan kaedah pengekstrakan pelarut separa-selanjar untuk
menyingkirkan molekul templat.
Kajian terhadap penjerapan terbutalina dengan pelbagai jenis pelarut telah
dijalankan untuk mengenalpasti pelarut-pelarut yang sesuai untuk digunakan semasa
penambahan sampel, pembilasan selektif dan pengelusian analit. Kajian tentang
selektiviti MIP berbanding polimer kawalan mendapati kedua-dua polimer menunjukkan
ciri pengenalpastian yang berbeza terhadap terbutalina. Graf pengelusian yang amat
berlainan menandakan kehadiran ikatan spesifik dalam MIP manakala pada polimer
kawalan pula hanya wujud ikatan yang lemah dan tidak spesifik. Kajian spesifikasi
silang pula menunjukkan MIP yang dihasilkan boleh mengenalpasti molekul-molekul
lain yang mempunyai struktur dan ciri-ciri yang hampir seiras dengan molekul templat.
Sampel urin dan plasma dicairkan dengan menggunakan 100 mM ammonium acetat
xxv
pH 7 pada nisbah 1: 1 v/v sebelum ia dialirkan melalui turus padatan yang diisi dengan
50 mg polimer. Kemudian turus padatan dibilas dengan 2 mL 50 mM ammonium acetat
pH 7 dan 1 mL asetonitril-asid asetik pekat (99: 1 % v/v). Pengelusian dilakukan
dengan menggunakan 4 mL asetonitril-asid asetik pekat (90: 10 % v/v). Proses
penerbitan dijalankan sebelum sampel dianalisis menggunakan GC-MS.
Hasil analisis sampel plasma memberikan 60 hingga 70 % pemerolehan
semula. Kajian pengesahan terhadap teknik yang digunakan mendapati ia memberikan
kepersisan dan ketepatan yang baik pada kepekatan antara 1 ng/mL hingga 10 ng/mL.
Teknik ini juga menunjukkan korelasi yang baik (r = 0.9860) dengan kaedah rujukan
yang berasaskan SPE fasa campuran. Ekstrak yang diperolehi adalah lebih bersih.
Kromatogram dan spektrum jisim yang dihasilkan juga kurang kompleks. Untuk kajian
pengenalpastian urin pula, kriteria yang ditetapkan oleh Jawatankuasa Olimpik
Antarabangsa (IOC) telah dipatuhi sepenuhnya. Perbezaan masa retensi bagi sampel
adalah kurang daripada 1 % berbanding masa retensi piawai manakala nisbah ion
yang dikaji menunjukkan had perbezaan maksimum kurang daripada 5 % (nilai mutlak)
dan 20 % (nilai relatif). Oleh itu, MIP merupakan di antara alternatif sebagai bahan
padatan SPE untuk pembersihan sampel biologi.
xxvi
SYNTHESIS AND APPLICATION OF MOLECULARLY IMPRINTED SOLID-PHASE
EXTRACTION FOR THE DETERMINATION OF TERBUTALINE IN BIOLOGICAL
MATRICES
ABSTRACT
A study was performed to evaluate the feasibility of applying MIPs as sorbent
material in SPE for clean up of terbutaline from urine and plasma samples. The
imprinted polymer was prepared by using methacrylic acid (MAA) as functional
monomer, ethylene glycol dimethacrylate (EGDMA) as cross-linker and salbutamol, a
closely structural analogue to the targeted analyte as template molecule. Porogen used
was a mixture of acetonitrile-methanol (75: 25 % v/v). Free radical polymerisation was
conducted at 50 oC for 24 hours. Molar ratio of template molecule: functional monomer:
cross-linker applied was fixed at 1: 4: 20. The output was a monolith macroporous
polymer which required grinding and sieving to obtain fine particles between 40 to 75
μm. Soxhlet extraction method was conducted to remove the imprint molecule in order
to create the recognition sites. A blank/non-imprinted polymer was produced
simultaneously using the same procedure except in the absent of template molecule.
A binding study of terbutaline in several solvents was performed to determine
suitable solvent for loading, selective washing and elution steps. Selectivity test on the
MIP against blank polymer demonstrated that both polymers exhibited different
recognition properties towards terbutaline. Significant differences in elution curves
between both polymers were observed, indicating the presence of specific binding in
imprinted polymer. In blank polymer, only weak non-specific interactions occurred.
Cross-specificity studies showed the MIP also exhibited molecular recognition
properties towards other structurally related compounds. Spiked urine and plasma
samples were diluted in 100 mM ammonium acetate buffer pH 7 (ratio = 1: 1 v/v) prior
to direct loading (5 mL) into a cartridge filled with 50 mg of MIP. Thereafter, the column
xxvii
was washed with 2 mL of 50 mM ammonium acetate buffer pH 7 followed by 1 mL of
ACN-glacial acetic acid (99: 1 % v/v). Finally, terbutaline was eluted with 4 mL of ACN-
glacial acetic acid (90: 10 % v/v). All samples were derivatised to form the trimethylsilyl
derivatives prior to GC-MS analysis.
Validation results of spiked plasma samples demonstrated good linearity,
specificity, precision and accuracy in the concentrations range of 1 to 10 ng/mL. A
recovery of 60 to 70 % was obtained. The method also correlated well (r = 0.9860) with
the reference SPE mixed mode method in plasma samples. Besides that, this method
also offers cleaner extract by obtaining less complex chromatogram and mass
spectrum. The International Olympic Committee (IOC) criteria for compound
identification were fulfilled for the spiked urine samples. Results obtained showed that
difference in retention times were less than 1 % between the samples and standards.
All ion ratios investigated have maximum permitted difference of less than 5 %
(absolute) and 20 % (relative). Therefore, MIP offers a competitive alternative as SPE
sorbent for effective clean up and enrichment of terbutaline in biological samples.
1
CHAPTER ONE
INTRODUCTION TO MOLECULARLY IMPRINTED POLYMER (MIP)
1.1. Introduction
Molecular imprinting is an established method for the production of polymeric
artificial receptors for specific molecular recognition. The imprinted polymers have the
ability of precise recognition of the original imprinting molecules and can distinguish the
minor structural differences of substrates in the interaction sites (Lu et al., 2002). Thus,
much of the literature available on the subject frequently underlines the “biomimetic”
properties by these imprinted polymers with the substrate-selective mechanisms being
analogous to that of natural entities such as antibodies and enzymes (Haupt et al.,
1998). Because of the high selectivity and stability of molecularly imprinted polymers
(MIPs), this technology has developed rapidly in recent years and is documented in
several comprehensive reviews (Mahony et al., 2005; Masque et al., 2001; Mosbach,
1994; Takeuchi et al., 1999b; Xu et al., 2004; Ye et al., 2004 and Komiyama et al.,
2003). In all the examples, MIPs have been used as substitutes for antibodies, showing
strong binding to the targeted analytes in affinity separations, assay systems and as
biosensors. Among the advantages of MIPs for analytical applications compared to the
conventional antibodies are easy preparation, chemical and thermal stability (Svenson
et al., 2001) and cost effectiveness, as the materials used are inexpensive and readily
available (Mahony et al., 2005). A summary of advantages and disadvantages of MIPs
is given in Table 1.1.
2
Table 1.1: Summary of advantages and disadvantages of MIPs for analytical applications (Mahony et al., 2005).
Advantages
Disadvantages
• Cost effective alternative to biomolecule-
based recognition
• Ease of preparation, enhanced thermal
and chemical stability versus natural
antibodies
• Can be prepared in different formats
(bead, block or thin film) following the
need of application
• Can be stored for a long period without
loss of affinity for target analyte
• Lower catalytic capabilities than
biological counterparts
• Unfavorable adsorption isotherm and
slow mass transfer in the polymer matrix
• Template bleeding requires suitable
template analogue for the imprinting step
and this will affects the quantitative
applications
• Grinding and sieving of bulk polymer for
SPE/LC application is labor intensive and
inefficient in material yield (high losses).
1.2. General Principle of Molecular Imprinting
The concept of MIP involves three main steps which are as follows (Komiyama
et al., 2003 and Sellergren, 2001):
• First step: Complex formation of a given template molecule with polymerisable
monomers bearing functional groups capable of interacting to each other by
covalent or non-covalent bonding. For non-covalent bonding, the functional
monomer and template are placed nearby through hydrogen bond, electrostatic,
hydrophobic, apolar or other non-covalent interactions.
• Second step: Polymerisation in order to maintain the alignment of the functional
groups which are optimally set for binding the template molecule. Structures of
the conjugates or adducts are frozen in a three dimensional network of polymer.
3
• Third step: Removal of the template molecule from the resulting polymer
matrices, allowing “tailor made” binding sites for the targeted analyte to be
generated. Hence, the space in the polymer originally occupied by the template
molecule is left as a cavity. Under appropriate conditions, these cavities
satisfactorily remember the size, structure and other physicochemical properties
of the template and bind this molecule or other structurally analogue molecule
efficiently and selectively.
Figure 1.1: Schematic diagram of molecularly imprinted polymer. I: self assembly of template
and functional monomers (non-covalent approach); II: polymerisation process; III: extraction of template and rebinding of analyte (Adapted from: Turiel et al., 2004).
4
1.3. Molecular Imprinting Approaches
Depending to the chemical bonding involved in molecular imprinting, the
technique can be classified into two systems, whether they are covalent bonding based
or non-covalent bonding based (Takeuchi et al., 1999b; Turiel et al., 2004 and
Komiyama et al., 2003).
1.3.1. Covalent Imprinting
The idea of covalent imprinting was first introduced by Wulff and co-workers in
1972 from Germany (Wulff et al., 1973 and Wulff, 1995) and followed by Shea and co-
workers from California (Mosbach, 1994). In the covalent system, a template-monomer
complex is formed through reversible covalent binding such as boronic acid esters,
acetals, ketals, Schiff bases, disulfide bonds, coordination bonds and others
(Komiyama et al., 2003). These linkages must be stable, reversible and for binding the
target guest promptly, both the formation and dissociation must be fast.
A typical covalent imprinting was boronic acid esters which are synthesised
from boronic acid and cis-1,2- or cis-1,3-diol compounds. After polymerisation, these
linkages are cleaved by hydrolysis and the boronic groups in the conjugates are
arranged suitably for guest binding. These conjugates are especially useful for
molecular imprinting towards carbohydrates and their derivatives which have cis-diol
moieties. Examples of this imprinting were discussed by Wulff et al., 1991 and Wulff et
al., 1997. For acetals and ketals bonding, ketone and aldehyde compounds are reacted
with 1,3-diol compounds and the resultant ketals and acetals products are used as
functional monomers. Imprinting with Schiff bases involved the reaction of aldehyde
with amino compounds to yield Schiff bases compounds. Coordination bonds involved
the interaction of metals ion with functional monomer to produce polymerisable metal
complexes, in which it acts as the functional monomers in the presence of an
5
appropriate ligand (template). Example of this type of imprinting was discussed by
Matsui et al., 1996.
1.3.2. Non-covalent Imprinting
Mosbach and co-workers have introduced this system to molecular imprinting
(Mosbach, 1994 and Komiyama et al., 2003). Compared to covalent imprinting, typical
interactions in non-covalent imprinting are hydrogen bonding, ionic, electrostatic, π-π
interactions, etc. Functional monomers are simply combined with template in the
polymerisation mixture and copolymerised with cross-linking agent. Procedure is
simple and easy to perform because it does not need to synthesise covalent
conjugates prior to polymerisation. Furthermore, the template can easily be removed
under mild conditions by simple extraction. However, non-covalent bonding may not be
strong enough to maintain template-functional monomer complexes. Thus, excess of
functional monomers are usually added to the reaction mixture in order to complete the
template-monomer complexation and to maintain stability under polymerisation
conditions. This results in a heterogeneous property of the binding sites in term of
affinity.
Many of the particular important molecules, example in pharmaceutical,
herbicides, biologically active substances and environmental contaminants possess
polar groups such as hydroxyl, carbonyl, amino and amide which are suitable groups
for non-covalent interactions. Of these, hydrogen bonding is the most appropriate for
precise molecular recognition since the bonding is highly dependent on both the
distance and direction between monomers and templates. However, for electrostatic
interactions, strong acids and bases are unfavorable as it is less dependent on
distance and direction. A proton is completely transferred to the base from acid. If the
combination consists of an intermediate strength of acid and base, hydrogen bonding
will be dominant and efficient imprinting should be achieved. Weak acids and bases
6
are also inappropriate for imprinting as the interactions are too weak. Because of its
simplicity and versatility, this technique has been widely attempted. For example,
Zurutuza et al., 2005 prepared a non-covalent molecularly imprinted solid-phase
material for the extraction of cocaine metabolites from aqueous samples. Vallano et al.,
2000 have successfully prepared a highly selective MIP column for capillary
electrochromatography. Other breakthroughs include the work from Zander et al., 1998;
Sellergren, 2001 and Kempe et al., 1994.
1.3.3. Comparison Between Covalent and Non-covalent Imprinting
There are significant differences between the covalent and non-covalent
imprinting. Table 1.2 shows the summary of advantages and disadvantages of these
two techniques (Komiyama et al., 2003 and Remcho et al., 1999). Among the
advantages of covalent imprinting are the monomer-template conjugates are stable
and stoichiometric. A wide variety of polymerisation conditions, example by high
temperature, at extreme pH and highly polar solvent can be employed since the
linkages are stable. For non-covalent technique, the main advantages are easy
removal of template molecule and the rate of guest binding and guest release is much
faster. Experimental work carried out throughout this study will focus on the non-
covalent imprinting technique as it is widely applied in various analytical fields.
7
Table 1.2: Advantages and disadvantages of covalent and non-covalent imprinting (Komiyama et al., 2003).
Factors Covalent Non-covalent
Synthesis of monomer-template conjugate
Necessary Unnecessary
Polymerisation condition Wide variety
Restricted
Removal of template after polymerisation
Difficult Easy
Guest binding and guest release
Slow Fast
Structure of guest binding site
Clearer Less clear
1.4. Synthesis of Molecularly Imprinted Polymer
To date, imprinted polymers in the form of particles are reportedly made by
various polymerisation methodologies, each of them developed to suit specific target
and application. So far, most of the MIPs have been prepared by bulk, suspension,
two-step swelling, precipitation and emulsion core-shell polymerisation (Perez-Moral et
al., 2004). Other less common methods include film synthesis, aerosol polymerisation
and polymerisation on silica particles. Each of these procedures involves the control of
different parameters during the synthesis and it produces polymers with different
properties and characteristics.
8
Figure 1.2: Schematical representation of the synthetic steps in the different polymerisation
procedures (Adapted from: Perez-Moral et al., 2004).
Table 1.3: Comparison of different polymerisation methods (Mayes et al., 1997).
Polymerisation Methods
Complexity Product Advantages Disadvantages
Bulk- in block Very straight forward
Random fragments after grinding
Simple, imprinting not affected by method derived factors
Tedious processing, wasteful, poor particle shape for HPLC
Bulk- in columns (in situ)
Moderate Solid block filling column
Simple, no column packing required
High back pressure, poor peak shape
Suspension – in water
Complex Spherical beads, polydisperse
Highly reproducible results, large scale possible, high quality beads
Water is incompatible with most imprinting procedures, only possible for some covalent and metal chelate based processes
Two-step swelling
Most complex
Monodisperse beads
Monodisperse beads, excellent packing for HPLC
Need for aqueous emulsion, rules out many imprinting processes
9
1.4.1. Polymerisation Method
1.4.1.1. Bulk Polymerisation
Chronologically, the first polymerisation method employed to synthesise a
MIP was based on the bulk polymerisation method. It is most widely applied by groups
working on imprinting because of its simplicity and universality (Mayes et al., 1997).
Basically all the components, which are mainly the template molecule, monomer,
cross-linker, initiator and porogen are mixed well and proceed to polymerise under
heating or ultra violet radiation. The result is a macroporous monolith polymeric block
that needs to be crushed and ground in order to obtain particles of irregular shape and
size. This process is time consuming and wasteful since a lot of the polymer is lost in
the process of grinding and sedimentation to eliminate fine particles. It may also
produce areas of heterogeneity in the polymeric matrix resulting from the lack of control
during polymerisation process, particularly when UV initiation is used.
1.4.1.2. Suspension Polymerisation
Suspension polymerisation in fluorocarbon solvent was first described by
Mayes and Mosbach (Mayes et al., 1997). It is a fast and reliable methodology that
synthesizes particles by ultra violet irradiation in less than 2 hour. The beads obtained
have a diameter that can vary between 5 to 50 μm depending on the stirring speed and
amount of surfactant added. It uses a perfluorocarbon or known as perfluoro-(1,3-
dimethylcyclohexane) solvent in the continuous phase which allows the same
interactions that occur in bulk polymerisation. The fluorocarbon suspending medium
can be easily recycled by distillation. This method offers a simple one step route to
high quality beads polymers with quantitatively yield of product and offers a very
attractive alternative to grinding and sieving, especially to chromatographic applications
10
(Mayes et al., 1997). When coupled with the use of trimethylpropane trimethacrylate
(TRIM) as cross linker, Mayes and group produced beads with high load capacities and
good separation.
1.4.1.3. Precipitation Polymerisation
Precipitation polymerisation is another method that can provides particles in
the submicron scale (0.3 to 10 μm). It is based on the precipitation of the polymeric
chains out of the solvent in the form of particles as they grow more and more insoluble
in an organic condition medium. There is no need of extra stabilizer because the
particles are prevented from coalescence by the rigidity obtained from the cross linking
of the polymer. The group of Cacho et al., 2004 successfully produced a polymer with
more homogeneous binding sites distribution and high affinity constants than those
obtained by bulk polymerisation using two different propazine imprinted methacrylic
based polymers as models. A non-covalent ferunon imprinted polymer which leaded to
the synthesis of spherical particles (∼ 1 μm) with homogeneous binding sites
distribution have been developed by Tamayo et al., 2003. They have successfully
evaluated the polymers for trace-enrichment and clean up of fenuron from plant sample
extracts.
1.4.1.4. Two-step Swelling Polymerisation
Two-step swelling polymerisation was developed with MIP by Hosoya and
Haginaka teams (Haginaka et al., 2002; Haginaka et al., 2004b; Haginaka et al., 2004a
and Hosoya et al., 1996). This technique requires several swelling steps on the initial
particles with the imprinting mixture before polymerisation proceeds. In this case, the
polymerisation medium is water. This method produces monodisperse particles in the
micron size range (2 to 50 μm) with good control of the final size and number of the
11
particles. It could be easily prepared, in-situ modification could be performed, and the
obtained MIPs are suitable for HPLC packing materials or SPE materials (Haginaka,
2004). A molecularly imprinted uniform sized polymer based stationary phase for
naproxen has been successfully developed by Haginaka et al., 1997. The imprinted
polymer materials showed enantioselectivity towards naproxen whereas the blank
polymer showed no chiral recognition ability. In another work of Haginaka et al., 2002,
they have obtained a uniformed enantioselective MIP for d-chlorpheniramine using
methacrylic acid and ethylene glycol dimethacrylate as functional monomer and cross-
linker respectively.
1.4.1.5. Emulsion Core-shell Polymerisation
This method produced core-shell particles in which they have a structured
morphology that allows the incorporation of any added property into the core of the
particles without interfering the imprinted shell (Perez-Moral et al., 2004). The
continuous medium during polymerisation is water. Particles obtained by this method
are monodisperse and can be used for surface imprinting. Generally, the seed was first
prepared using a standard batch emulsion polymerisation in a three-necked jacketed
reactor connected to a water bath to control the temperature. The system was
equipped with a condenser, a mechanical stirrer and a gas inlet to maintain an inert
argon atmosphere. A solution of sodium hydrogen carbonate and sodium dodecyl
sulphate in distilled water was added to the reactor and purged with argon to remove
oxygen under gentle stirring, while increasing the temperature to 90 oC. Once the
temperature was reached, the monomer mixture (methacrylic acid and ethylene glycol
dimethacrylate) was introduced into the reactor and the stirring speed increased to 600
rpm. Then, the initiator was added to initiate the polymerisation. The temperature was
maintained at 90 oC for 24 hours and the final latex was filtered through a fine nylon
mesh. Next, the core-shell particles were synthesised using the similar method as for
12
preparing the seed. Solution containing water and sodium dodecyl sulphate was added
into the reactor and purged with argon. Gentle stirring was applied and solution of
monomer, cross-linker, template molecule and seed that have been previously mixed
was charged into the reactor followed by an aqueous solution of ammonium
peroxodisulphate. The stirring speed was increased and the reaction is allowed to
proceed for 6 hours before cooling to room temperature. Non-imprinted polymer was
prepared following the same procedure but excluding the template from the formulation.
1.4.2. Polymerisation Reagents
The essential chemicals required to produce a MIP are the functional monomer,
template molecule, cross-linker, initiator and polymerisation solvent or known as
porogen. Chemical structures of the typical reagents used are as presented in Figure
1.3.
O
OH
N
O
NH2
O
O
CH 3
O
O
CH 3
Methacrylic Acid 4-Vinylpyridine Acrylamide
Ethylene Glycol Dimethacrylate (EGDMA)
Functional Monomer
Cross-linking Agent
Figure 1.3: Typical reagents for polymerisation.
13
H3C
CN
N N
CH3
CH3
CNCH3
CH3
CH3
CH3
H3C
CN
N NCH3
CH3
CN
Azobisisobutyronitrile (AIBN) Azobisdimethylvaleronitrile (ABDV)
Radical Initiator
Figure 1.3: Typical reagents for polymerisation (continued).
1.4.2.1. Template Molecule
In all imprinting process, template is one of the most important components.
The template chosen must be chemically inert and stable under polymerisation
conditions since all polymerisations are based on the free radical interactions. The
template molecule must not participate in the radical reaction and stable upon
exposure to UV or high polymerisation temperature (Cormack et al., 2004). Usually, a
closely structural analogue to the targeted analyte was chosen as template molecule.
This is to prevent the template leaching or bleeding problem during analysis especially
for quantitative analysis at trace level as not the entire template molecules are
successfully extracted out from the imprinted polymer even after extensive washing
(Martin et al., 2004). Blomgren et al., 2002 used brombuterol as template for the
analysis of clenbuterol in calf urine. The structures of these two molecules are very
similar and the MIP obtained was selective and sensitive for analysis of clenbuterol at
low level. Martin et al., 2000 also employed the similar approach by preparing
propanolol imprinted polymer to extract five compounds structurally related to
propanolol.
14
1.4.2.2. Functional Monomer
Functional monomers are responsible for the binding interactions in the
imprinted binding sites. Normally, in non-covalent imprinting, it was added in excess
relative to the number of template molecule. The ratio of template to functional
monomer of 1: 4 and upwards are rather common (Cormack et al., 2004). It is very
important to match the functionality of the template with the functionality of the
monomer in a complementary fashion in order to obtain maximum complex formation.
Basically, functional monomers can be divided into three different groups which are
mainly the acidic, basic and neutral monomer. Methacrylic acid is the most common
acidic monomer and are widely selected by various groups of researchers (Dong et al.,
2004; Zurutuza et al., 2005 and Theodoridis et al., 2002). For acidic template
molecules, vinylpyridine and acrylamide can be selected as functional monomer (Zhou,
et al., 1999; Simon et al., 2004 and Bastide et al., 2005).
1.4.2.3. Cross-linking Agent
A cross-linker is added to fulfill three major functions in imprinting. The
fundamental role is to fix and control the morphology of the polymer matrix besides
from stabilise the imprinted binding sites. They also make the imprinted polymer
insoluble in solvents and impart the mechanical stability to the polymer matrix. High
cross-link ratios are generally preferred in order to access permanently porous
materials and to generate materials with adequate mechanical stability. Thus, polymers
with a high degree of cross-linking (70 to 90 %) are required (Masque et al., 2001; Cai,
et al., 2004 and Komiyama et al., 2003). For effective imprinting, the reactivity of cross-
linker should be similar to that of the functional monomer in a cocktail polymerisation to
ensure smooth incorporation of the co-monomer. Several of well-known, commercially
15
available cross-linking agents such as ethylene glycol dimethacrylate (EGDMA) and
divinylbenzene (DVB) are compatible with molecular imprinting.
1.4.2.4. Porogen
The function of polymerisation solvent is to bring all the components into one
phase. Besides that, it also responsible for creating pores in macroporous polymer.
The nature and level of porogen added will determine the morphology and total pore
volume. Thermodynamically good solvents tend to lead to polymers with well
developed pore structures and high specific surface areas whereas thermodynamically
poor solvents lead to poorly developed pore structures and low specific surface areas.
Selection of the porogen is mainly dependent on the type of imprinting. For covalent
imprinting, many kind of solvents are employed as long as they are able to dissolve all
components. However in non-covalent imprinting, it is critical to the formation of
conjugates between the functional monomer and template. Normally, this implies that
apolar and non-protic solvents such as acetonitrile, toluene and chloroform are
preferred as such solvents stabilise the hydrogen bonds (Chassaing et al., 2004 and
Mena et al., 2002). However, in certain application, polar protic porogen was chosen
even though polar solvent will disrupt the hydrogen bonding (Baggiani et al., 2001 and
Caro et al., 2004).
1.4.2.5. Initiator
The initiator is added into the system to initiate free radical polymerisation.
When the initiator is trigged either by heat, UV radiation or chemicals, carbon centered
free radicals will formed and these unpair electrons are capable to react with other
monomer /cross-linker in order to propagate into longer chains. Polymerisation process
stopped when two free radicals reacted with each other (Young et al., 1992). The
16
amount of initiator added is relatively at low level compared to the monomer. Azo
initiator such as azobisisobutyronitrile can be conveniently decomposed by photolysis
(UV) or thermolysis.
1.4.3. Factors to Consider in the Synthesis of Selective MIP
1.4.3.1. Molar Ratio of Template: Monomer: Cross-linker (T: M: X)
Number and quality of the MIP recognition sites are highly dependable on the
molar relationship between template and functional monomer. The common optimum
mole ratio of template molecule, monomer and cross-linker for production of MIP is 1:
3-5: 20-30 (Komiyama et al., 2003). Theoretically, high molar ratio of T: M affords less
than optimal complexation on account of insufficient functional monomer and too low of
T: M causes non-selective binding (Andersson et al., 1999). Results by Andersson et
al., 1999 clearly indicate that an excess of either template or functional monomer
during polymerisation is unfavourable regard to selectivity. The group prepared a series
of polymers with different T: M ratios for selectivity test. Polymer with the ratio of T: M =
1: 4 has the best selectivity properties as compared to the others. Baggiani et al., 2004
research supported the finding as polymers prepared at T: M = 1: 15 and 1: 20
exhibited poor recognition effect as it is difficult to clearly discriminate them from the
corresponding blank polymers. Experiments carried out by Theodoridis et al., 2004
showed that despite high molar ratio of T: M, high affinity recognition sites would be
limited as the agglomeration of template in organic solvent environment could occur.
Thus, polymers prepared at the ratio of 1: 2.7: 13.4 exhibited poor recognition
properties compared to polymers synthesised at ratio 1: 46: 230 and 1: 4.6: 23. Other
groups such as Caro et al., 2004; Mena et al., 2002; Spivak et al., 2001 and Zander et
al., 1998 also synthesised their imprinted polymers according to the ratio of T: M: X = 1:
4: 20. However, the work done by Davies et al., 2004 showed that their optimum
17
predicted ratio for T: M: X were 1: 10: 55 and 1: 10: 10 according to the chemometrics
approach. Several examples of reagents selection for polymerisation process and the
molar ratio of T: M: X employed by various research groups are described in Table 1.4.
1.4.3.2. Stability of Monomer-template Assemblies
A stable monomer-template assembly is also important for achieving a larger
number of imprinting sites and at the same time, the number of non-specific binding
sites will be minimised (Sellergren, 1999). Therefore, the type of functional monomer
selected and porogen choices are very important to produce a stable monomer-
template conjugate. As mentioned previously in Section 1.4.2.2 and 1.4.2.4, monomer
selected should be able to serve as a hydrogen bond, proton donor and as a hydrogen
bond acceptor. Best porogen will be the aprotic and non polar solvents as these
solvents have poor hydrogen binding capacity and low dielectric constant. Thus, they
lead to large interaction energy between the template and the functional monomer,
resulting in a better affinity and selectivity. Work from Schmidt et al., 2005 and Wu et
al., 2005 have proven that the influences of porogens are essential on the affinity and
selectivity of MIPs. Figure 1.4 shows the factors affecting the recognition properties of
MIPs related to the monomer-template assemblies.
References
Description
Ersoz, A. et. al., 2004
Lu, Y. et. al., 2004 Svenson, J. et. al., 2001
Cacho, C. et. al., 2004
Zhou, J. et. al., 1999
Template molecule (T)
4- nitrophenol L-phenylalanyl-aminopyridine
Theophylline Propazine 5,5-diphenylhydantoin
Monomer (M) Cross-linker (X) Porogen Initiator
Methacrylamido-antipyrine EGDMA Acetonitrile Azobisiso-butyronitrile
Methacrylic acid EGDMA Chloroform Azobisiso-butyronitrile
Methacrylic acid EGDMA Chloroform Azobisiso-butyronitrile
Methacrylic acid EGDMA Toluene Azobismethyl-propionitrile
Acrylamide
18
Table 1.4: Choice of reagents and molar ratio of T: M: X for the synthesis of MIPs by bulk polymerization technique.
EGDMA Tetrahydrofuran Azobisiso-butyronitrile
Ratio of T: M: X
∼ 1: 6: 29 ∼ 1: 4: 21 ∼ 1: 4: 18 ∼ 1: 4: 20 ∼ 1: 4: 20
Initiation method
UV radiation Thermal Thermal Thermal
Polymerization process
60 oC for 24 hours 60 oC for 24 hours
25 oC for 24 hours 65 oC for 20 hours
Refer to landscape table in the folder
19
Figure 1.4: Factors affecting the recognition properties of MIPs related to the monomer-template
assemblies (Adapted from: Sellergren, 1999).
1.4.3.3. Polymerisation Temperature
Temperature is an important factor influencing the recognition properties of
MIPs as it affects the polymerisation process and polymer structure. The polymer’s
affinity and specificity can be improved significantly by optimising the polymerisation
temperature (Lu et al., 2004). According to the research done by Lu and co-workers,
lower polymerisation temperature is advantageous to the stability of the template-
functional monomer assemblies in the pre-polymerisation mixture. However, higher
polymerisation temperature is favorable for completeness of the polymerisation
reaction, which improves the quality and quantity of MIPs recognition sites. They
prepared three types of polymers imprinted with 3-L-phenylalanylaminopyridine at 10
20
oC, 40 oC and 60 oC for 24 hours respectively. Polymer prepared at 40 oC has both the
highest enantioselectivity and largest sample load capacity as compared to polymers
prepared at 10 oC and 60 oC. Besides that, study conducted by Piletsky et al., 2002
suggested that the polymer is able to memorise the temperature used in the
polymerisation process in a manner similar to previously documented MIP memory
effects for the template and polymerisation solvent.
1.5. Polymer Structure and Morphology
The structure integrity of the monomer-template assemblies must be preserved
during the polymerisation process to allow the functional groups to be confined in
space in a stable arrangement complementary to the template. However, the role of the
polymer matrix is not only to contain the binding sites in a stable form but also to
provide porosity allowing easy access for the guest to all sites. This can be achieved by
applying a high level of cross-linking agent and sufficient porogen during
polymerisation. Thus, most of the cross-linked network polymers used for imprinting
have a wide distribution of pore sizes associated with various degrees of diffusional
mass transfer limitations and a different degree of swelling.
1.5.1. Types of Pores
Pores sites in an imprinted polymer can be classified according to different
types based on the site accessibility, integrity and stability criteria. Nitrogen adsorption-
desorption and mercury porosimetry are techniques for the determination of polymer
pore structures in a dry state (Sellergren, 1993). There are typically three types of
pores which are the mesopores, macropores and micropores (Sellergren, 1999). Meso
and macropores with pore size larger than 20 Å are expected to be easily accessible
compared to sites located in the smaller micropores (pore sizes smaller than 20 Å)
21
where the diffusion is slow. The number of the latter may be higher since the surface
area for a given pore volume of micropores is higher than that of macropores. For most
applications in liquid media, permanent porosity and a large surface area of accessible
meso and macropores are preferred. Referring to Figure 1.5, this gives materials
containing mainly accessible sites of type A and B although a significant number of
non-specific sites (type F) may present. One of the undesirable effects of adding an
excess of template is the loss of site integrity due to coalescence of the binding sites
(type D), which is related to the extent of template self-association. Site G contributes
to problem of extracting template molecule as the template is remain strongly bound to
the polymer even after careful extraction.
Site A: In macropores Site B: In micropores Site C: Embedded Site D: Site coalescence Site E: Induced binding site Site F: Non-specific site Site G: Residual template
Figure 1.5: Types of binding sites in MIPs (Adapted from: Sellergren, 1999).
1.5.2. Adsorption / Binding Isotherms
Adsorption or binding isotherms can yield important information concerning the
binding energies, modes of binding and sites distributions in the interaction of small
molecule ligands with receptor as happened in MIPs. In MIPs, a soluble ligand interacts
with binding sites in a solid adsorbent. The adsorption isotherms are plots of
22
equilibrium concentrations of bound ligand (adsorbate) versus concentration of free
ligand. This isotherm helps to characterize the MIPs and calculate the corresponding
binding parameters and affinity distributions. Adsorption or binding isotherms for MIPs
can be obtained from batch rebinding studies in which a constant weight of polymer is
equilibrated with a known concentration of analyte (Umpleby et al., 2004). This is then
measured over a range of analyte concentrations. The concentration of the analyte
remaining free in solution is measured by HPLC, UV spectroscopy or radio-ligand
assay. The corresponding concentration of bound analyte is calculated as the
difference between the total and free concentrations. Selection of binding model is
primarily based on its ability to accurately reproduce the experimental isotherm. The
physical basis for the model should also reflect the distribution of sites found in the
measured system in order to generate realistic binding parameters.
The isotherm can be fitted using various models where different assumptions
are made. Generally, the models can be grouped into two classes which are the
discrete and continuous distribution models (Umpleby et al., 2004). Langmuir and bi-
Langmuir isotherms are the most commonly applied of discrete binding models. These
models simplify a distribution into a finite number of different classes of sites, with each
class of site having a different binding affinity. The Langmuir model assumes there is
only a single class of sites and the bi-Langmuir assumes there are only two classes of
sites. The Freundlich and Langmuir-Freundlich are both examples of continuous
distribution models in which a continuous function containing an infinite number of
different types of binding sites is used to model the distribution. These models provide
more accurate approximations for the heterogeneity present in most MIPs and also
provide quantitative measures of heterogeneity. Studies conducted by Umpleby and
co-workers suggested that MIPs contain a broad unimodal distribution that
exponentially tails into the high affinity region. This exponentially decaying region
appears to be the most important with respect to the enhanced affinity and selectivity of
MIPs.
23
1.6. Application of Molecularly Imprinted Polymers
In analytical separation science, molecularly imprinted polymers have been
applied in several analytical techniques such as in liquid chromatography, capillary
electrochromatography and capillary electrophoresis, solid-phase extraction,
immunoassay and as a selective sorbent in chemical sensors. The improved selectivity
of imprinted polymers compared to conventional sorbents may lead to cleaner
chromatographic traces in the subsequent analytical separation.
1.6.1. Affinity Based Solid-phase Extraction
Out of all the MIPs applications, the most widely acceptance and close to
practical realisation is probably that of solid-phase extraction (Andersson, 2000). A
number of groups have presented SPE applications based on MIP for various types of
analytes in various types of samples matrices. The matrices include biological fluids,
tissues, food, aqueous environment samples and pharmaceutical compounds. As
shown by Zander et al., 1998, analysis of nicotine and its oxidation products in nicotine
chewing gum was possible by applying MIPSPE technique. Besides from obtaining
high and reproducible recoveries of less polar analytes, this MIP method eliminates
liquid-liquid extraction steps which consume lots of organic solvents. Research by
Mena et al., 2002; Caro et al., 2004; Gallego-Gallegos et al., 2005 and Zurutuza et al.,
2005 have successfully synthesised and evaluated MIP as selective SPE sorbent for
the analysis of aqueous samples. Other examples are extracts of various analytes from
plasma and serum samples (Andersson et al., 2004; Bereczki et al., 2001; Theodoridis
et al., 2002; Wu et al., 2004 and Mullett et al., 1998). Chassaing et al., 2004 have
developed MIP packed into 96-well blocks enabling high throughput analysis and good
accuracy and precision. Their research showed that analysis with MIP produced much
cleaner baseline, leading to lower background noise and higher sensitivity as
24
compared to the conventional C18 SPE method. Table 1.5 summarises the application
of MIP as SPE sorbent. As the technology is becoming well known, there are
commercially available MIPSPE in the market. One of the leading companies is MIP
Technologies AB from Lund, Sweden (www.miptechnologies.se). This company offers
various ranges of chromatography products from SPE to HPLC. Studies conducted by
Kootstra et al., 2005 and Widstrand et al., 2004 using MIPSPE columns from MIP
Technologies for multi-residues clean-up of beta-agonists in bovine muscle and calves
urine have proven that the method meets the requirements for quantitative
determination.
Table 1.5: Summary of studies in which MIPs have been applied to SPE.
Analyte
Sample Application Reference
Cocaine metabolites (Benzoylecgonine)
Aqueous samples Selectively extracted and quantified at clinical relevant concentration.
Zurutuza et al., 2005
Naphthalene sulfonates
Environment samples
Analysed water from Ebro river.
Caro et al., 2004
Tributyltin Seawater Quantitation limit of 0.04 μg/L for 1 L.
Gallego-Gallegos et al., 2005
Pirimicarb Water samples On-line pre-concentration. Mena et al 2002
Phenytoin Plasma Good linearity at 2.5 – 40 μg/mL
Bereczki et al., 2001
Local anesthetics (bupivacaine, ropivacaine and mepivacaine)
Human plasma Determination of bupivacaine: 3.9 – 500 nmol/L Deternmination of ropivacaine: 7.8 – 500 nmol/L
Andersson et al., 2004
Esculetin Ash bark of traditional chinese medicine
- Hu et al., 2005b
Atrazine Beef liver extracts Tolerance level in meat products: 0.02 ppm
Muldoon et al., 1997
Theophylline Serum - Mullett et al., 1998
Monosulfuron Soil
Recoveries (MIPSPE-HPLC): > 93 %, CV < 3.2 %
Dong et al., 2004
Cephalexin and α-aminocephalosporin antibiotics
Human serum Linearity of Cephalexin: 0.3 – 25 μg/mL Analysis using MIPSPE-PE-MS
Wu et al., 2004